Emerging technology is coming at us thick and fast. IoT, augmented reality, neural networks, big data, robotic process automation, deep learning, analytics, wearables, blockchain, bots, and digitalisation. This field is becoming so complex that even for the best CIOs, CDOs and CTOs in the industry trying to understand and explain the role of emerging technology is overwhelming. Here’s an attempt to tie it all together and explain it in plain english.
In a parallel industry scientists have been searching for a universal theory of everything to help explain how all the latest concepts in quantum physics and general relativity fit together. While that will no doubt be of immeasurable benefit to humankind, what we as ICT leaders really need is a universal theory of digitalisation. A universal theory of digitalisation would provide us a simple and technology agnostic way of explaining how digitalisation, and all the numerous emerging technologies associated with it, can improve our lives.
Thanks to futurists such as Gartner who have a very successful track record of predicting the future of technology, we are no longer wondering what technology is coming, the new challenge is what to do with the technology when it gets here? In this short article, I will show you a way of explaining the digital ecosystem in a way that hopefully makes sense to everyone.
I’m going use a story about a pilot we have underway in New Zealand to help improve the safety of people working in dangerous situations. Parts of the story are underway right now, some of it is still at the vision stage, and some of it is embellished for the sake of helping explain the concept, but all of it is possible today.
[For those that prefer video to reading, I recently gave a talk at the IoT Wellington user group which covered this article. Someone was kind enough to film it and post it, so if you want to see and hear the concept rather than reading about it here’s the video. Sorry about the quality.]
What is Digital?
What does the term ‘digital’ mean? It has become the catch phrase of this decade, much like e-business was in the ‘90s, when we started to have e-everything, e-mail, e-business, e-commerce, e-marketing and e-service. By the end of that decade doing things electronically was the norm so we got over the ‘e’ and went back to calling marketing just good old marketing. Then the 2000’s dotcom decade began and we swapped out all the e’s for i’s and did it all again, i-startups, i-business, i-marketing and i-commerce. Now welcome to the decade of digital-business, digital-marketing and digital-assets. At least this time we haven’t started calling it d-business (I hope I didn’t just start something there).
The best way to think about the term ‘digital’ isn’t as a single technology or a thing, but as an ecosystem of technologies that all form part of a process, I call it the D.I.G.I.T.A.L Ecosystem.
- Data – gather data points
- Information – what is happening?
- Gaps – what don’t we know?
- Insight – why is it happening?
- Think – what can we do to change it?
- Act – make the change
- Learn – did it work?
This process can be used to solve problems, enhance a customer experience, change society, or in our example, help save lives, improve worker safety, and improve productivity.
This is example is based around an organisation that must send its people into extremely hazardous situations, involving extreme heights and energy that could kill in an instant. When you are responsible for the lives of people in those situations human safety is your number one priority, making sure your people get home to their families every day becomes an obsession.
So how do we use the D.I.G.I.T.A.L Ecosystem to address this challenge? Let’s walk through the seven steps.
We need to gather some data points. The employees in the risky situations are given an intelligent biometric wearable, a Fitbit on steroids if you like. This gathers a wide range of vital signs which allows us to know if they are showing signs of stress, have they fallen off a ladder, have they been electrocuted, have they stopped moving, or are they just too tired.
This is of course the Internet of Things (#IoT). There’s the first part of the digital puzzle slotted into place.
Next we move to Information. Raw data is pretty meaningless to a human, it needs to be presented in the form of information, preferably in simple graphical dashboards. So, we gather all the data and present it in a dashboard to show the real-time status of the people, and the historical trends of their wellbeing. Now we can see what is happening. For the first time the organisation can see the wellbeing of their staff in real-time.
Now we’ve added Business Intelligence (#BI) and reporting to the story.
Now that we can see what is happening some very concerning things come to light. At certain times of the week a large part of the afterhours workforce simultaneously display symptoms of extreme stress.
IS it an overload of electricity, a brownout, a nasty email from the boss? The biometric information we have only gives us a narrow view of the world, the employee’s status, but we don’t know what is going on around them. We have a gap, we need more data.
This is where #bigdata comes in – by linking in public data such as weather feeds, power demand, traffic congestion, and social media, and then adding in the organisation’s data such as emails, calendar appointments, equipment status and job sheets. This gives us a wider view of the world so we can look for correlations.
Now we have a vast amount of data, a #data lake, but unfortunately we have too much data for a human to understand. People can track correlations between a few data sets, but when we don’t know what is causing the problem we don’t know which data sets to look at. Most correlations between data are too distant for mere humans to see, and with so much data our ability to see anything decreases in relation to the amount of data we have.
Now we feed the data into a #machine learning system. We don’t need to build one, there are plenty of cloud based systems, be it AWS, Google or Azure, they all have the ability to find correlations between events that don’t seem obvious to us.
Now we have Insight – we can see the status of the people, and what events are happening in the world around them, now we can use #data visualization to show connections between the events.
Insight can enable us to eliminate false positives. Turns out a spike in stress symptoms every Wednesday at 8:30pm for 16 weeks of the year might not be anything to do with a failure in the power grid. Thanks to the social media data we have in the data lake we can see that it just happens to coincide with the airing of the latest season of Game of Thrones.
But seriously, insight gives us the ability to see what the leading causes of harm are, and then we can start to address them.
Think – now we have vast amounts of data, we have visualisations that can show us the connections between different events, we can start to see what is causing the most harm to our people.
The question is, what is the best way to reduce harm? It’s not a simple question because we must balance worker safety with productivity, economics, and the availability of power to customers. We could have a 100% safe workforce if they never left the office, but we’d be out of business and the country would be in the dark, literally.
With all that data, we now have thousands of levers to pull that could make a change, but which ones, and how far?
So now we use artificial intelligence (#AI) to help learn the most efficient solution to the problem. AI can model the data and outcomes, and come up with the optimal way to reduce harm to our people, while still enabling us to provide a reliable energy supply at an economical cost.
Time to Act. Now we know what to do, we need to change a few processes. Rather than assigning the nearest technician to a fault, we need to assign the one that is best able to deal with the complexity and risk. A technician that has been on the job for seven hours and is showing signs of fatigue shouldn’t be sent into a highly dangerous situation. We need to connect our job dispatch system to the AI system and use AI to help match the right people to the job.
Sadly, the systems that run the job dispatch process are legacy with no APIs, so how do we get an intelligent system to interface with them? Traditionally we could use a human to copy the data from one screen and type it into another screen, but its slow, its expensive and its highly prone to human error. It’s not the way to manage high risk job dispatch to keep a critical utility online.
Instead we can use robotic processes automation (#RPA), the robot software takes data from one form, validates it, analyses it, makes an informed decision and enters it into another form. In this case, it robotically reads the fault data from one legacy system, passes it to the AI, and then robotically types it back into a job dispatch system. Fast, efficient and 100% accurate.
This optimises the job allocation process, ensuring the right mix of workers’ skill, mental state, and physical state to reduce as much harm as possible.
Learn – finally we loop back to the start and use the IoT, BI, BigData, Machine Learning, and Visualisation to show us if the changes are having a positive impact. Then the AI and robotic process automation can learn and adjust to continually improve.
There in one short story (all be it fictional in parts) is an explanation of how the D.I.G.I.T.A.L process joins together all of today’s emerging technology to solve problems and improve outcomes. Now when you are asked “what role does [insert new technology here] play in our business?” you have a way of explaining it in non-technical terms.
I’ve shown how the #D.I.G.I.T.A.L process can be used to integrate and explain all of the new technologies available to us, and I’ve shown that you don’t start with the technology, you start with the problems you face, and the outcomes you want to achieve.